How to choose the right KPIs: fewer indicators, better decisions.

When a company shows us its indicator dashboard, the first warning sign is the quantity. Forty charts in a monthly report do not mean much is being measured; they mean nobody decided what matters.

An indicator deserves a place on the management dashboard only if it meets four conditions. First, someone can act on it: if the result changes no decision, it is context, not a management indicator. Second, it has a named owner, not a department. Third, a target exists before measuring, not adjusted afterwards to fit the result. Fourth, the data source is single and auditable.

The most frequent mistake is confusing activity with results. "Number of training sessions delivered" measures effort; "percentage of leaders who applied the tool within 60 days" measures results. "Open recruitment processes" measures motion; "time to fill critical vacancies" measures the organization's real capacity to operate.

The second useful distinction is between outcome and process indicators. The first say what happened: quarterly turnover, personnel cost over revenue. The second anticipate what is going to happen: percentage of performance conversations held, vacation days accumulated above the limit. A dashboard with only outcome indicators informs late; one with only process indicators loses sight of the goal.

Each indicator needs three elements to be actionable: the figure, a comparison — against target, against the previous period or against the market — and the explanation of the variance. Without the third, the dashboard informs but does not allow decisions, and the monthly meeting is spent finding out why a number went up instead of fixing it.

Our practical recommendation for a general manager: between six and ten indicators, reviewed monthly, with one person responsible for the figure and another for the explanation. Everything else lives in each area's operational dashboards, where detail does make sense.